Reproduced from Ian Rye's 3 November 2020 LinkedIn article. Read the original on LinkedIn →
A prototype, by its very nature, is going to be defined very differently based on what it is a prototype of. Prototype software differs wildly from a prototype sketch or a prototype machine. Even beyond that, there are a variety of prototype categories: proof-of-principle, working, visual, user experience. In this series we'll specifically be discussing an automotive production-intent prototype — whether or not it can be "perfect," and what that even means.
"Perfect Prototype" is probably a poor name for what we're talking about. What we're really discussing is a Perfect Prototype Build: the goals of an entire build phase — when an automotive company builds a fleet of production-level vehicles for the sake of testing — going as well as possible. The focus is to succeed in probably three specific areas:
- The prototype build costs the least amount of money possible (spending no money is unfeasible, so despite the word "perfect," we're really shooting for "minimised").
- The prototype assembly is started as soon as possible, and the build and subsequent full testing phase are completed as soon as possible.
- The maximum number of potential design improvements and issues for the final product are identified, evaluated, and actioned (if decided upon).
Is this possible? In reality, no. If you've ordered 232 clips when only 231 were necessary, you've removed your shot at a Perfect Prototype Build. It is, however, very possible to get extremely close.
Where to target for savings
Let's talk about some of the seemingly obvious targets. Designing and engineering the parts takes the longest and costs a ton. So if we finish designing earlier, we start building earlier, right? It's not a bad thought — but the truth is that there will always be plenty of bottlenecks in that stage of the game that are going to keep you from cutting it down too much. Engineering can only really start once you've locked down your entire concept, so if you figure out what you want to build earlier then everything would move earlier.
It's also just easy to target engineering. If we think of wasted time as a sort of equal coefficient across all business activities, then 5% of saved time in engineering goes a lot further than 5% time saved in kitting parts or building the vehicles. But what if we can save 30% of the sourcing time? What if we can save 50% of the manufacturing cost? After you've pulled most of the releasing onus from engineers and pushed them back into the world of engineering (the prime directive that even birthed QR_ in the first place), it may be more valuable to target other time- and cost-saving areas in the pursuit of Perfect Prototype Build goals.
Take the first goal. What are the excess costs of a prototype build that really drain the budget? Too many parts ordered. Incorrect parts ordered. Correct parts that aren't the correct tooling pedigree or design maturity. Unnecessary vehicles built. Unnecessary parts ordered for vehicles that don't require them. The list is endless — and since "time is money" applies more heavily to the mass manufacturing industry than maybe anywhere else in the world, the second goal of starting and completing the prototype phase as soon as possible leaks into the first. If I've started my prototype build a day late, my forecasted first sold vehicle is now back a day later than it could be — and even a single day's worth of sales can be enormous.
There's potentially even more to be saved within the testing itself. Are my tests streamlined? Can I use the same vehicle built for testing door stability after 10,000 slams to then test how the suspension holds up after 100,000 miles of Michigan potholes? Is this test even necessary? Am I forgetting some tests that should be necessary? And well before that, how am I managing my failure modes and effects analysis (FMEAs)? Are the right parts being analysed? Are too many parts being analysed? Is it being done soon enough to optimise when the testing plan can be put into place?
I want to tackle some of the main areas where we can drive efficiency into the system. I hope to give a clear, wide-scope view of the main sections of the process that can see improvement — and a granular, detailed look at how some individual PLM decisions can drastically alter an automotive company's success in the prototype world. Specifically, I want to answer:
- How can we improve communication and drive action-transfer between sections of the company to improve time to market?
- How can we gain an earlier view on the prioritisation of parts and actions to mitigate job-stoppers and slippage?
- Is there a way to know what needs to be tested sooner, to give more time to prepare for the build?
- How much weight does a Bill of Materials validation hold on the overall success of the prototype phase?
Hopefully some of the messages here make it into your OEM or other workplace — and we can continue to deliver better products earlier and for a more affordable price.
